Search for: All records

Most current strategies for carbon management require CO2 removal (CDR) from the atmosphere on the multi-hundred gigatonne (Gt) scale by 2100. Mg-rich silicate minerals can remove >105 Gt CO2 and sequester it as stable and innocuous carbonate minerals or dissolved bicarbonate ions. However, the reaction rates of these minerals under ambient conditions are far too slow for practical use. Here we show that CaCO3 and CaSO4 react quantitatively with diverse Mg-rich silicates (for example, olivine, serpentine and augite) under thermochemical conditions to form Ca2SiO4 and MgO. On exposure to ambient air under wet conditions, Ca2SiO4 is converted to CaCO3 and silicic acid, and MgO is partially converted into a Mg carbonate within weeks, whereas the input Mg silicate shows no reactivity over 6 months. Alternatively, Ca2SiO4 and MgO can be completely carbonated to CaCO3 and Mg(HCO3)2 under 1 atm CO2 at ambient temperature within hours. Using CaCO3 as the Ca source, this chemistry enables a CDR process in which the output Ca2SiO4/MgO material is used to remove CO2 from air or soil and the CO2 process emissions are sequestered. Analysis of the energy requirements indicates that this process could require less than 1 MWh per tonne CO2 removed, approximately half the energy of CO2 capture with leading direct air capture technologies. The chemistry described here could unlock Mg-rich silicates as a vast resource for safe and permanent CDR. 

Abstract Geodetic strain rate characterizes present-day crustal deformation and therefore may be used as a spatial predictor for earthquakes. However, the reported correlation between strain rates and seismicity varies significantly in different places. Here, we systematically study the correlation between strain rate, seismicity, and seismic moment in six regions representing typical plate boundary zones, diffuse plate boundary regions, and continental interiors. We quantify the strain rate–seismicity correlation using a method similar to the Molchan error diagram and area skill scores. We find that the correlation between strain rate and seismicity varies with different tectonic settings that can be characterized by the mean strain rates. Strong correlations are found in typical plate boundary zones where strain rates are high and concentrated at major fault zones, whereas poor or no correlations are found in stable continental interiors with low strain rates. The correlation between strain rate and seismicity is also time dependent: It is stronger in seismically active periods but weaker during periods of relative quiescence. These temporal variations can be useful for hazard assessment.